Is the hype around software defined networking subsiding? Research firm Gartner seems to think so. The July 2015 edition of the organization's vaunted "hype cycle" chart plots the evolution of different technological trends from an innovation trigger to a peak of inflated expectations to a trough of disillusionment to a slope of enlightenment and finally to a plateau of productivity. The idea is that technologies are initially marketed to an overly broad population - many segments of which will never find any real, tangible use case for them - only to fall out of the public eye as a smaller group of actual adopters begins moving in. From there, they mature.
For example, a recent formulation of the hype cycle put virtual switches on the plateau of productivity, 802.11ac Wi-Fi on the slope of enlightenment and SDN in the trough of disillusionment. Gartner's analysts figured that SDN was still five to 10 years out from maturity. This stance makes sense, since SDN is typically something that will be phased in over many years, as an evolutionary rather than revolutionary strategy, as Gartner's David Cappuccio explained at a summit in May 2015.
What's the next step for SDN?
The end of SDN hype does mean the end of SDN's prospects - quite the opposite, in fact. As it becomes better understood by service providers, there will more opportunities to weave it into their networks according to realistic multi-year plans. This is exactly what is happening with initiatives such as AT&T's ambitious Domain 2.0.
"SDN applications are still at the height of the hype curve."
SDN applications, created by writing to northbound and southbound APIs, are still at the top of Gartner's hype curve, so it could be a bit longer before they come into their own. The rise of open source platforms for SDN such as the Open Networking Operating Systems (backed by Huawei, Intel, NTT Communications and many others) may help push SDN to a more mature phase.
"By now, SDN is deployed in data centers worldwide, based on proprietary software," said Scott Shenker, faculty director of the Open Network Research Center. "The next frontier for SDN is service provider networks, where large network operators need to program their networks to create new, differentiated services. To enable this, we need a highly available, scalable control plane such as ONOS upon which new services can be instantiated and deployed."
The takeaway: SDN hype is dying down. At the same time, actual work on SDN implementation, rather than just marketing, is underway. Projects such as ONOS could pave the way for more effective SDN orchestration in the years ahead.